# Agent 1 — Lead Finder Agent

**Build Priority:** 1 (Highest)
**Time to Build:** 2–3 days
**Impact:** ⭐⭐⭐⭐⭐ — No leads, no business
**Cost/Month:** $50–100 (AirDNA)

## What It Does

Finds STR-viable properties owned by absentee owners in your target market. Replaces hours of manual tax record hunting.

## Inputs

- Target zip codes and counties
- STR legality map (which zones allow short-term rentals)
- Property criteria (bedrooms, home type, price range)

## Outputs

- A validated CSV of property addresses + owner names + owner mailing addresses
- Enriched with AirDNA estimated monthly revenue
- Scored by "likelihood to convert" (absentee + STR-legal + high revenue potential)

## Data Sources

| Source | How to Access | What You Get |
|--------|--------------|--------------|
| County Assessor | Scrape or buy via PropStream ($99/mo) or Reonomy | Owner name, mailing address, property details |
| AirDNA | AirDNA API or MarketMinder export ($50–100/mo) | Comparable STR revenue, occupancy, ADR |
| Zoning/STR Map | Scrape municipal STR registry or buy Host Compliance data | STR-legal yes/no per address |
| Google Maps API | Geocoding + neighborhood data | Proximity to landmarks, transit, demand drivers |

## Tech Stack

- **Data Pipeline:** Python (pandas) + cron job (daily)
- **APIs:** PropStream REST API / county assessor CSV downloads
- **Enrichment:** AirDNA bulk export → merge on address
- **Scoring model:** Simple weighted (absentee +30, STR-legal +40, revenue >$3k +20, 2+ bedrooms +10)
- **Output:** CSV → Google Sheets (via Google Sheets API)

## How to Build

```bash
mkdir lead-finder-agent && cd lead-finder-agent
python -m venv venv && source venv/bin/activate
pip install pandas requests google-auth google-api-python-client
```

Core class structure:
```python
class LeadFinderAgent:
    def __init__(self, counties, zip_codes): ...
    def pull_tax_records(self): ...
    def filter_absentee_owners(self, df): ...
    def enrich_with_airdna(self, df): ...
    def check_str_legality(self, df): ...
    def score_leads(self, df): ...
    def push_to_sheets(self, df): ...
```

## Deployment

**Schedule:** Once per week (Monday morning)
**Platform:** Hermes cron job or any cron scheduler
**Trigger:** Manual run or scheduled weekly
